Is That a Stimulus Package in Your Pocket?: The Impact of the American Recovery and Reinvestment Act on Aggregate Demand

Abstract

Due to the collapse of the housing bubble and the financial crisis, the US aggregate demand curve shifted leftward, causing the Great Recession of 2008-2009. In response to the recession, the $832 billion American Recovery and Reinvestment Act of 2009 (ARRA), otherwise known as “the stimulus,” was passed to push the aggregate demand curve outward and mitigate the recession. The ARRA consisted of policies such as tax cuts, public investment outlays, and aid to state and local governments. The ARRA was controversial because prior to the recession, it was commonly accepted that monetary policy was sufficient to push the aggregate demand curve outwards in response to negative demand shocks. The ARRA was also criticized because state and local governments allegedly did not increase their government purchases in response to ARRA funds. Finally, the public criticized the ARRA because the Council of Economic Advisors had previously estimated that without the ARRA, unemployment would skyrocket to about 9%. However, with the ARRA’s passage, unemployment went up to 10%, leading many to label the ARRA as a failure. The purpose of this thesis is to evaluate the ARRA by estimating its effects on aggregate demand. To do this, this thesis exploits state-level variation in ARRA funding to estimate the effects that the ARRA had on two major components of aggregate demand: personal consumption expenditures and state and local government purchases. The parameter of interest is the multiplier, which measures the change in aggregate demand induced by a change in ARRA funding. To account for the concern that variation in ARRA funding may be endogenous to states’ post-recessionary economic conditions, I instrument for ARRA funding, with my instrumental variables coming from either the political economy of federal spending or ARRA funding formulas that should be orthogonal to states’ postrecessionary economic conditions. Instrumental variable regressions generally lead to positive and statistically significant effects from the stimulus on personal consumption expenditures and on state and local government purchases, suggesting that the ARRA shifted the aggregate demand curve outward, as intended. However, the ARRA seems to have stronger and faster effects on personal consumption expenditures than on state and local government purchases. Furthermore, the ARRA also has positive and statistically significant effects on states’ rainy day funds, implying that state governments saved some of their ARRA funds. The policy implication of this thesis is that state and local governments are suboptimal agents for countercyclical fiscal policy, and that it would be more effective for fiscal policy to target low-income individuals who have a high marginal propensity to consume. Furthermore, there is substantial variation in the size of multipliers, with the size of the multiplier depending on the federal agency allocating the ARRA funds. However, despite the fact that the ARRA had a stronger effect on personal consumption expenditures than on state and local government purchases, the ARRA has clear and positive effects on state and local government purchases, suggesting that the criticism that state and local governments did not increase their purchases in response to the ARRA is inaccurate.